530 Chronicles of Science. [July, 



beds. 3rd. The metamorphism of the latter into quartzite. 4th. The 

 denudation of the overlying strata. 5th. The denudation of the 

 quartzites during the Triassic period, and the formation of the pebble- 

 bed. 6th. The denudation of the pebble bed during supracretaceous 

 periods, by which much of the ancient gravel of south-eastern Devon- 

 shire was produced. 7th. The continued denudation of the pebble- 

 bed to form the modern beach of Budleigh Salterton. The author 

 thinks that the crystalline rocks of the south of Devon and Cornwall, 

 including those of the Bolt and Start, the Dodman district, and Eddy- 

 stone, may be parts of one great system, from which the Budleigh 

 Salterton pebbles may have been derived. 



The Introduction of Cavern Accumulations. — In this paper Mr. 

 Pengelly shows that the caverns of Devonshire may be divided into 

 four classes, respecting the modes in which their contents were intro- 

 duced. In some instances, at least, the existing contents of caverns 

 were carried in long after the excavation had been completed. The 

 famous Windmill Hill Cavern at Brixham had certainly been twice 

 filled and emptied before the introduction of the mass of materials 

 which was found in it and excavated in 1858. This is proved by the 

 remains of two former stalagmitic floors, one of which formed a kind 

 of ceiling to a cavity which existed when the cavern was examined, and 

 below which was the most recent accumulation of materials. 



The first mode in which accumulations are shown to be introduced 

 into caverns is that through vertical openings communicating with 

 the surface. Several caves occurring in the limestone of Oreston 

 were met with from 1816 to 1822, and described by Mr. Whidbey, 

 who failed to detect any such opening. But in 1858, Mr. Pengelly 

 had an opportunity of examining one of the same series, situated, as 

 he believed, along the same original line of fracture. The roof was 

 eight feet thick, and consisted of large angular masses of limestone 

 cemented with carbonate of lime. This was evidently the position of 

 the original fissure, through which the materials of the bone bed had 

 been slowly and gradually introduced. There had been pauses in the 

 operation, as shown by the frequent occurrence of pure white stalag- 

 mite separating masses of breccia. The breccia forming the roof 

 might easily be mistaken for ordinary limestone traversed by numer- 

 ous irregular and coarse veins. Other examples are mentioned of 

 fissures filled from above, and destitute of roof, though containing 

 bones of the ordinary cave animals well preserved. In some instances, 

 as at Daddy's Hole, Torquay, the fissure has been only in part filled 



Another mode of introduction is by engulphed streams. The 

 Brixham Cavern consists of a series of galleries having the direction 

 of the two systems of joints which pervade the Devonian rocks of the 

 district, but their roof-joints are too close-fitting to allow anything 

 more than water to penetrate. In this case, however, the cavern has 

 four external horizontal entrances, and the mode of introduction is 

 strikingly shown by the fact that every bone and stone lay with its 

 longest axis in the plane of the deposit, or in the precise position 

 which would have been given to it by a running stream, and not by 



